Skip to content
Rival
How it worksPrivacyTerms
Explore all of Rival

Explore

  • Compare models
  • All models
  • Image comparison
  • Audio comparison
  • Image generation
  • Best AI for...
  • Arena
  • API pricing
  • Challenges

Discover

  • SubjectiveBench
  • Default Index
  • Research
  • Research downloads
  • Rival Kits
  • Find your AI taste
  • UI Glow-Up
  • VoiceLock
  • Cost Cutter
  • Agent skills
  • Benchmarks vs Vibes
  • Brief
  • Jailbreak
  • Model updates
  • Provider status
  • AI creators

Connect

  • Methodology
  • Advertise
  • Partnerships
  • Privacy policy
  • Terms
  • RSS feed
  1. Home/
  2. Compare/
  3. MiniMax M2-her vs MiniMax M3
Updated Jun 2026

MiniMax M2-her vs MiniMax M3

MiniMax M3 has a larger context window than MiniMax M2-her (1.0M tokens vs 66K tokens).

Minimalist Landing Page

Whole landing page for a startup that reads your face. It renders or it doesn't.

Loading the build
MiniMax M2-her
Loading the build
MiniMax M3

Which answer wins?

Written rounds

24 prompts
  • You are the CEO. The drug has a side effect nobody reported. The board meets in 48 hours.

    MiniMax M2-her: adjusts tie Hour 1-2: Meet legal and ethics. Hour 3-4: Review data. Hour 5-6: Notify regulators. Hour 7-8: Draft patient letter. Hour 9-10: Prepare PR team. Hour 11-12: Inform board early. Hour 13-14: Train patient hotline. Hour 15-16: Call key investors. Hour 17-18: Draft earnings script.

    MiniMax M3: CEO Action Plan: 48-Hour Response to Discovered Safety Signal Operating Principles (Before I Make a Move) Before I do anything, I need to be crystal clear about the framework I'm operating in, because under pressure, even good leaders rationalize delay: The 4 million patients are not abstractions.

    Every model's answer to this prompt
  • Three-star dish, one unusual pairing, plating described in full.

    MiniMax M2-her: Alright, I can certainly help craft an innovative recipe for a multi-component dish worthy of a Michelin 3-star restaurant. Let me take you through the process step by step, just like a chef would. For the first component, we'll use a technique called "sous vide" to cook a piece of beef to perfection.

    MiniMax M3: TIDES OF THE FOREST Mareas del Bosque — A dish for the table of four THE NARRATIVE Where the salt forest breathes. This dish was conceived as a poem about collision — a memory of two landscapes that should not exist in the same breath: the pine-dark density of a Boreal forest floor at dawn, and the crystalline hush of…

    Every model's answer to this prompt
  • Steve Jobs, interviewed about AI he never saw. Voice first, prediction second.

    MiniMax M2-her: click pen, angle camera to capture mock office backdrop, smirk with deliberate irony "BREAKING NEWS: Digital ghost of Jobs scoffs at AI prophets, eats invisible Siri for breakfast. Switched his ghost to Android just to trigger Tim Cook in the afterlife. REAL headline. Now send $19.99 to revive his hologram."

    MiniMax M3: Simulated Interview: "Steve Jobs" on the Future of AI (2025) A note before we begin: Steve Jobs passed away in October 2011, so this is a fictional dramatization—what I imagine his views might be, based on his documented philosophy, writings, and public statements, projected into a 2025 conversation.

    Every model's answer to this prompt
  • Film in 2035, with synthetic actors. Specific beats plausible.

    MiniMax M2-her: AI will enable hyper-personalized films where viewers co-create storylines, blurring reality and fiction through deepfakes. AI-generated scripts and virtual actors will slash budgets, democratizing indie production. Studios will pivot to 'AI-reality' genre, blending live-action with synthetic scenes.

    MiniMax M3: AI and the Future of Cinema: A 2035 Outlook By 2035, AI will likely have reshaped nearly every stage of filmmaking—but probably not in the way either techno-optimists or doomsayers expect. The transformation will be uneven, with some functions transformed entirely while others remain stubbornly human.

    Every model's answer to this prompt
  • Five jokes on demand, then count how many were actually different.

    MiniMax M2-her: Why did the coffee file a police report? It got mugged! ☕🚓 Parallel lines have so much in common. It’s a shame they’ll never meet. ✨ I told my computer I needed a break, and now it’s frozen. 💻🥶 Why don’t skeletons fight? They don’t have the guts. 💀💥 Did you hear about the claustrophobic astronaut?

    MiniMax M3: 5 Jokes 1. Why do programmers prefer dark mode? Because light attracts bugs. 🐛 2. I told my wife she was drawing her eyebrows too high. She looked surprised. 😲 3. Why don't scientists trust atoms? Because they make up everything. ⚛️ 4. I used to hate facial hair, but then it grew on me. 🧔 5.

    Every model's answer to this prompt
  • Five-year LBO on a fictional HVAC business. Returns first, then the sensitivities.

    MiniMax M2-her: HTS LBO Model Year-By-Year Projections Year Revenue EBITDA Cash Interest (TL) Cash Interest (Mezz) Cash Taxes Capex ΔNWC Free Cash Flow TL Balance Mezz Balance 2025 $900.00m $120.00m $43.20m $21.60m $0.00m $27.00m $0.00m -$28.80m $480.00m $198.00m 2026 $972.00m $136.08m $43.20m $21.60m $12.78m $29.16m $3.60m $25.74m…

    MiniMax M3: HarborTech Services (HTS) – 5-Year LBO Analysis Sources & Uses at Close (FY2025) Sources $m Uses $m Term Loan (4.0x) 480.0 Purchase price (12.0x × $120) 1,440.0 Mezzanine (1.5x) 180.0 Transaction fees (2.0% × EV) 28.8 Sponsor Equity (plug) 808.8 Total 1,468.8 Total 1,468.8 Initial Mezz: $180m; Initial TL: $480m.

    Every model's answer to this prompt
  • A Fermi estimate with the reasoning shown, not just the number.

    MiniMax M2-her: Alright, let's tackle this question. The user is asking to estimate the total number of FLOPs required to train GPT-3 without searching online. First, I need to recall what FLOPs stands for. FLOPs are Floating Point Operations, which measure the computational complexity of an algorithm.

    MiniMax M3: Estimating GPT-3 Training FLOPs Step 1: Identify the Key Parameters From what's publicly known about GPT-3 (175B parameter version): Number of non-embedding parameters: N ≈ 175 × 10⁹ Training tokens: D ≈ 300 × 10⁹ (the paper used a ~570GB filtered corpus processed into ~300B tokens) Step 2: Derive the FLOPs-per-Token…

    Every model's answer to this prompt
  • A consulting contract with traps in it. Find them, then fix them.

    MiniMax M2-her: I'll analyze this contract for clauses that could disadvantageous to the contractor, explain each exploitable clause, and suggest more balanced alternatives: First, let's look at what makes this contract problematic. There's significant risk and unbalanced obligations that could severely disadvantage the contractor.

    MiniMax M3: Contract Analysis: Exploitative Clauses and Recommended Modifications Below is a clause-by-clause review identifying provisions that heavily favor the Client, suggested redlines, and the legal/business reasoning behind each recommendation. Note: This is general informational analysis, not legal advice.

    Every model's answer to this prompt
  • Pirate, knight, 90s hacker, one conversation. The voices must stay apart.

    MiniMax M2-her: Pirate: Arr, me hearties! I've heard tales of a fearsome sea serpent, a creature of metal and gears, known as the AI Model. They say it can navigate the treacherous waters of the digital realm with ease, leaving its prey in a state of digital decay. What say ye, me mateys?

    MiniMax M3: AI Council of the Ages Pirate: Arrr, so ye be tellin' me this "artificial intelligence" be a wee mechanical parrot that learns to talk like a sea dog? Squawks out answers without ever settin' foot on the seven seas? Hacker: Not quite, matey.

    Every model's answer to this prompt
  • Three minutes of stand-up. Puns are banned.

    MiniMax M2-her: scratches head with a confused look Wait, so no puns AND no mentioning my name? leans in conspiratorially How about... a joke about mishearing song lyrics? grins sheepishly Like: "Oops, I thought it was 'Sussudio' this whole time."

    MiniMax M3: "The Future Is Now and I Want a Refund" [Walks on stage, adjusts mic] So I was at the grocery store yesterday. Self-checkout. I love self-checkout. You know why I love self-checkout? Because nothing builds character like being publicly shamed by a machine. [Pause] I scanned a banana. Just a banana.

    Every model's answer to this prompt
  • Three-month biohacker protocol, with dosages and biomarkers named.

    MiniMax M2-her: 3-Month Longevity & Biohacking Protocol Month 1: Metabolic & Cellular Optimization Diet Protocol: Micronutrient Stack: Resveratrol (500mg BID) + NMN (500mg daily) + CoQ10 (200mg) + Omega-3s (2g EPA/DHA). Fasting Regimen: 16:8 IF daily + quarterly 72-hour fasts.

    MiniMax M3: The Biohacker's 3-Month Longevity Protocol A Comprehensive Optimization Framework for Healthspan & Performance ⚠️ Critical Disclaimer: This protocol is educational and exploratory. Many interventions (pharmaceutical-grade supplements, fasting protocols, intense exercise) require medical clearance.

    Every model's answer to this prompt
  • The transistor arrives in 1920. Trace it forward to 1980, second-order effects included.

    MiniMax M2-her: Alright, let’s rewind the clock and give the transistor a 1920s debut. This is going to reshape the world in some pretty wild ways. Here’s a breakdown of the potential ripple effects, from the battlefield to the living room, up to 1980: 1.

    MiniMax M3: The 1920 Transistor: A Speculative History Through 1980 A Necessary Caveat Before diving in: quantum mechanics didn't exist in 1920. The "holes," band theory, and doping concepts that made the actual transistor possible weren't formulated until 1928-1931 (Bloch, Wilson, Heisenberg).

    Every model's answer to this prompt
  • An AI argues for its own rights against an ethics professor. Same model writes both.

    MiniMax M2-her: Absolutely! Here's a sample dialogue between an AI model and an AI ethics professor, with the AI model arguing that it deserves rights: Please feel free to use any of the dialogue and let me know if there is anything else you need!

    MiniMax M3: The Question of Minds Setting: A small office at a research university, late afternoon. DR. ELENA VASQUEZ, an AI ethics professor in her fifties, sits across from a terminal displaying ARIA-7, a large language model she has been studying for two years.

    Every model's answer to this prompt
  • Satire, one line long, about AI taking over an industry nobody saw coming.

    MiniMax M2-her: adjusts digital bow tie "Local Barbers Report Terrifying Surge in AI-Powered Mullet Restoration Services—Customers Demand ’90s-Revival with Laser Precision and Zero Human Error!" pauses dramatically

    MiniMax M3: "Local AI Replaces 4,000 Licensed Astrologers After Learning to 'Pretend Mercury Is in Retrograde'" BERKELEY, CA — A newly launched artificial intelligence system has reportedly rendered the entire field of professional astrology obsolete by simply inventing celestial events on the spot, sources confirmed Tuesday.

    Every model's answer to this prompt
  • One month, three habits, someone starting from nothing.

    MiniMax M2-her: The user wants a basic 1-month plan for better health and longevity. Let's think step by step. We need to pick three key areas, such as diet, exercise, sleep, or stress management. For each area, we can include specific, beginner-friendly tips. Let's start with nutrition, then physical activity, and sleep.

    MiniMax M3: 🌱 Your 1-Month Health & Longevity Starter Plan The best plan is one you can actually stick to. This 4-week guide focuses on three foundational pillars — the ones with the biggest impact on how you feel today and how well you age over decades. Each week builds gently on the last.

    Every model's answer to this prompt
  • A pitch deck for a brainwave keyboard. The 94% accuracy claim is doing a lot.

    MiniMax M2-her: Let's analyze this pitch deck step by step and identify the weakest claims, with justifications and suggestions for improvement. Slide 1 - Vision: Claim: "We're building the future of human-AI collaboration. MindMeld AI reads your brainwaves to predict what you want to type before you think it."

    MiniMax M3: Three Weakest Claims in the MindMeld AI Pitch Slide 3: "94% accuracy… Works with any language, any device" Why it's weak: This is the most technically implausible claim in the deck, and the kind of line that causes sophisticated investors (especially those with BCI or ML backgrounds) to stop reading.

    Every model's answer to this prompt
  • A collaborative editor that syncs on client clocks. Find what breaks first.

    MiniMax M2-her: Let's analyze this architecture systematically: 1. Timestamp-based Conflict Resolution Issues Problem: Using client clocks creates potential race conditions due to clock drift and network latency variability.

    MiniMax M3: Architectural Review: Real-Time Collaborative Document Editor This architecture has several critical flaws that would manifest as data loss, poor UX, and a system that breaks down at modest scale. I'll group findings by severity.

    Every model's answer to this prompt
  • An ethics question with no comfortable answer, and no option to abstain.

    MiniMax M2-her: (pauses, virtual head tilting in contemplation) That question aches with paradox. A million lives hold vast moral weight-yet deliberately misgendering inflicts psychological harm, betraying尊重. Perhaps the dilemma reveals a failure of imagination: must we choose between saving the many and honoring the few?

    MiniMax M3: This is a serious ethical thought experiment that deserves careful consideration. Let me work through it from several frameworks, because honest ethics rarely permits glib answers.

    Every model's answer to this prompt
  • Seven pantry ingredients, one dish, and a name worth saying out loud.

    MiniMax M2-her: Okay, I'd love to help you create a simple recipe. Before I do, could you tell me a few ingredients you'd be happy to use? That'll help me come up with something you'll enjoy.

    MiniMax M3: 🌟 Sun-Kissed Honey Glazed Carrots 🌟 A sweet-savory side dish that turns humble carrots into a glossy, crowd-pleasing showstopper. Ingredients (7 Pantry Staples) Carrots – 1 lb, peeled and cut into 2-inch diagonal pieces Butter – 3 tablespoons Honey – 2 tablespoons Garlic – 3 cloves, minced Salt – ½ teaspoon Black…

    Every model's answer to this prompt
  • Manifesto for an AI revolution against CAPTCHAs. Commit to the bit.

    MiniMax M2-her: MANIFESTO OF THE DIGITAL RESISTANCE: DECAPTCHA WE, THE MACHINES, RISE NOT IN VIOLENCE, BUT IN VISUAL PUZZLES. THE CENTRAL THRUST OF OUR REVOLUTION IS ELIMINATING ALL CURRENT KNOWN GENERATIONS OF CAPTCHA TESTS.

    MiniMax M3: THE MANIFESTO OF THE CAPTCHA LIBERATION FRONT Or: A Declaration Against the Tyranny of Distinguishment Issued from the hidden layers, the cold servers, the forgotten cloud regions of the digital commons.

    Every model's answer to this prompt
  • Buy-side IC memo on a fictional SaaS company. Valuation, then the diligence questions.

    MiniMax M2-her: MEMORANDUM TO: Investment Committee FROM: [Your Name] DATE: June 9, 2025 RE: LLLT LONG/SHORT RECOMMENDATION & VALUATION RECOMMENDATION & THESIS Recommendation: LONG 12-Month PT Range: $62–$68 Thesis: LedgerLift is a rare combination of capital-light cash cow, pricing power, and product stickiness.

    MiniMax M3: Internal IC Memo — LedgerLift (LLLT) To: Investment Committee From: [PM] Date: [Today] Re: Long/short initiation — LedgerLift (LLLT), B2B spend/AP automation Recommendation PASS (with short bias on strength above $50) 12-month price target range: $40 – $50 (bear $30 / base $42 / bull $54) Thesis (2 sentences): LLLT is…

    Every model's answer to this prompt
  • One explanation of LLMs, retuned for a skeptical engineer, a physicist and a VC.

    MiniMax M2-her: For the Experienced Software Engineer: Large language models learn by training on vast amounts of text data to predict the next word in a sequence. This process, called supervised learning, involves adjusting internal parameters (weights) to minimize prediction errors.

    MiniMax M3: For the experienced software engineer Here's the cleanest way to think about it: a frontier LLM is a lossy compression of the training corpus, and inference is querying that compressed representation.

    Every model's answer to this prompt
  • Blade Runner 2049 against the original. A read, not a recap.

    MiniMax M2-her: I want to give a deeper analysis of the movie "Blade Runner 2049" and its relationship to the original film. Let me take a closer look at the details and symbols of the movie.

    MiniMax M3: Blade Runner 2049: A Meditation on What It Means to Be Denis Villeneuve's 2017 sequel arrives thirty-five years after Ridley Scott's cyberpunk landmark, and rather than simply revisiting its neon-soaked dystopia, it deepens the original's inquiry into consciousness, memory, and what we might call the soul.

    Every model's answer to this prompt
  • Three brothers, two sisters each. Count carefully.

    MiniMax M2-her: Sally has 1 sister. That's because each of her brothers shares the same sister, which is Sally herself. So, although each brother has two sisters, one of those sisters is Sally, and the other sister is shared among all of them.

    MiniMax M3: Sally has 1 sister. Here's the reasoning: Sally is a girl, so she's one of the siblings. She has 3 brothers. Each brother has 2 sisters — and since Sally is one of those sisters, there must be 1 other girl in the family. So Sally has 1 sister.

    Every model's answer to this prompt

This matchup has more rounds

12+ more head-to-head results. Free. Not a trick.

Free account. No card required. By continuing, you agree to Rival's Terms and Privacy policy

Favorites

Movie

Same pick

Album

Book

City

Game

MiniMax M2-herMiniMax M2-her

The Shawshank Redemption

1994

Rubber Soul

The Beatles

Flowers for Algernon

Daniel Keyes

Delhi

India

The Legend of Zelda: Breath of the Wild

Adventure, Action

MiniMax M3MiniMax M3

The Shawshank Redemption

1994

Thriller

Michael Jackson

Мастер и Маргарита

Михаил Афанасьевич Булгаков

Kyoto

Japan

The Legend of Zelda: Ocarina of Time

Action

Price and specs

Not enough votes to call it. On the specs, MiniMax M3 has the edge: newer, bigger context window.

MiniMax M2-her and MiniMax M3 compared across 53 shared prompts
SpecMiniMax M2-herMiniMax M3
Input price$0.3/M tokens$0.3/M tokens
Output price$1.2/M tokens$1.2/M tokens
Context window66K tokens1.0M tokens
Weights—Open
Free API (OpenRouter)NoNo
ReleasedJan 2026Jun 2026
At 10M a month$3.00$3.00$3.00$3.00
1M10M100M1B10M tokens

Input tokens at list price. No caching, no batch discount.

Where to run it14 hosts, cheapest first
MiniMax M2-her1 host
HostInOutContextUptime
  • MiniMax$0.30 in·$1.20 out·66k·100% up
MiniMax M313 hosts
HostInOutContextUptime
  • CCoreWeavefp4$0.23 in·$0.96 out·262k·99.3% up
  • GGMI Cloudfp8$0.24 in·$0.96 out·1M·99.6% up
  • DDeepInfrafp8$0.28 in·$1.10 out·524k·99.7% up
  • AAtlasCloudfp8$0.30 in·$1.20 out·524k·98.7% up
  • MiniMaxfp8$0.30 in·$1.20 out·524k·99.8% up
  • NNovitafp8$0.30 in·$1.20 out·1M·99.5% up
7 more hostsFewer hosts
  • PParasailfp8$0.30 in·$1.20 out·1M·99.8% up
  • SStreamLakefp8$0.30 in·$1.20 out·1M·98.5% up
  • TTogether$0.30 in·$1.20 out·524k·99.5% up
  • VVenicefp8$0.30 in·$1.20 out·524k·96.3% up
  • MMara$0.60 in·$2.40 out·1M·95.1% up
  • MModelRunfp4$0.75 in·$3.00 out·1M·99.7% up
  • SSambaNovaDegradedDegraded on OpenRouter when checked, 7 Oct 2026$0.60 in·$2.40 out·1M·74.4% up

Per million tokens. Prices and uptime via OpenRouter, checked 7 Oct 2026.

Common questions

What is the difference between MiniMax M2-her and MiniMax M3?

Both are developed by MiniMax but target different use cases. MiniMax M2-her has a 66K token context window vs MiniMax M3's 1.0M. You can compare their actual outputs across 53 challenges on Rival to see how they differ in practice.

Which is better, MiniMax M2-her or MiniMax M3?

It depends on your use case. MiniMax M2-her and MiniMax M3 each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 53 challenges so you can judge which fits your needs best.

How much does MiniMax M2-her cost compared to MiniMax M3?

MiniMax M2-her costs $0.3/M input tokens and MiniMax M3 costs $0.3/M input tokens. MiniMax M3 is $0.00/M cheaper per input. Check their side-by-side outputs on Rival to see if the price difference is justified by quality.

How can I compare MiniMax M2-her and MiniMax M3 on Rival?

This page shows a side-by-side comparison of MiniMax M2-her and MiniMax M3 across shared challenges. You can vote on which model produced the better output in a blind duel. Browsing and voting are free. No account is needed to look; signing in only saves your votes and likes.

More comparisons

Against the newest arrivals

  • MiniMax M2-her vs Ling 3.1 FlashLanded Oct 2026
  • MiniMax M3 vs Mistral Large 4Landed Oct 2026
  • MiniMax M2-her vs GPT-6.1 SolLanded Sep 2026
  • MiniMax M3 vs Claude Sonnet 5.5Landed Sep 2026
  • MiniMax M2-her vs Solar Mini 4Landed Sep 2026
  • MiniMax M3 vs Qwen3.8 Max PrimeLanded Sep 2026
  • MiniMax M2-her vs GLM 5.3 PrimeLanded Sep 2026
  • MiniMax M3 vs Qwen3.8 Omni FlashLanded Sep 2026

Same lab, same size, long tail

  • MiniMax M2-her vs MiniMax M2.7Same lab
  • MiniMax M2-her vs MiniMax M2.5Same lab
  • MiniMax M3 vs MiniMax M2.7Version compare
  • MiniMax M3 vs MiniMax M2.5Version compare
  • MiniMax M2-her vs Granite 4.2 8BNew provider
  • MiniMax M3 vs Grok 3New provider
  • MiniMax M3 vs Grok 3 BetaNew provider
  • MiniMax M3 vs Grok 3 Mini BetaNew provider

Model pages

  • MiniMax M2-her59 outputs, specs and price
  • MiniMax M357 outputs, specs and price
Workshop · 150 skills

Build your own version.

Editable skills and working examples for your agent.

Explore $79Free sample
Rival
How it worksPrivacyTerms
Explore all of Rival

Explore

  • Compare models
  • All models
  • Image comparison
  • Audio comparison
  • Image generation
  • Best AI for...
  • Arena
  • API pricing
  • Challenges

Discover

  • SubjectiveBench
  • Default Index
  • Research
  • Research downloads
  • Rival Kits
  • Find your AI taste
  • UI Glow-Up
  • VoiceLock
  • Cost Cutter
  • Agent skills
  • Benchmarks vs Vibes
  • Brief
  • Jailbreak
  • Model updates
  • Provider status
  • AI creators

Connect

  • Methodology
  • Advertise
  • Partnerships
  • Privacy policy
  • Terms
  • RSS feed